We evaluated the accuracy of the angle described by Levine and Drennan, the metaphyseal-diaphyseal angle of the proximal aspect of the tibia, for the differentiation of physiological bowing from Blount disease. We compared this angle, as measured at presentation, in 106 children (179 extremities) who had physiological bowing with the angle in nineteen children (thirty-two extremities) who had documented Blount disease. The angle averaged 9 +/- 3.9 degrees for the patients who had physiological bowing and 19 +/- 5.7 degrees for the patients who had Blount disease (p < 0.0000001). Linear regression analysis, performed to evaluate any changes in the metaphyseal-diaphyseal angle in relation to age, showed that the older the child was at the time of presentation the more likely it was that the angle would be smaller in a child who had physiological bowing and larger in a child who had Blount disease. The chance for false-positive and false-negative errors was greater than 5 per cent if the angle was more than 9 degrees but less than 16 degrees. Sixty-six (37 percent) of the 179 extremities in the group that had physiological bowing had an angle of at least 11 degrees; one extremity affected by Blount disease had an angle of less than 11 degrees. In this study, we found that the metaphyseal-diaphyseal angle may be helpful in the identification of Blount disease but should not be the sole criterion used to determine the diagnosis.